for the automated extraction of the information about the height and deformation velocity of single and multiplescatterers in the same radar cell. More specifically, the contribution of this paper is threefold. First, the originator method ofsuperresolution adaptive Diff-Tomo is improved beyond the first tests in  in order to handle automatically a large data set by augmenting it with a new original scatterer multiplicity detector for the extraction of the height/deformation velocity information. In particular, we tackle the detection problem by combining adap- tive Diff-Tomo with a model-based least squares (LS) fitting in the complex (i.e., amplitude and phase) data domain. All the theoretical details and the discussion about the novelty of the proposed identification (i.e., detection and parameter estimation) algorithm are reported in Section II. Second, by means of the developed scatterer multiplicity detector, the extracted height/deformation velocity information is validated extensively, instead ofon a number of selected cells. Third, a preliminary phenomenological analysis of the characteristics of the detected scatterers is performed, which is novel also considering the superresolutionof the adaptive processing. These two points and the related results have been addressed in Section III by processing real C-band ERS-1/2 data over an area around the San Paolo stadium in the city of Naples and over the Cinecittà area in the city of Rome, with particular emphasis on single and double scatterers.
Abstract—Synthetic aperture radar tomography (TomoSAR) extends the synthetic aperture principle into the elevation direc- tion for 3-D imaging. It uses stacks of several acquisitions from slightly different viewing angles (the elevation aperture) to re- construct the reflectivity function along the elevation direction by means of spectral analysis for every azimuth–range pixel. The new class of meter-resolution spaceborneSAR systems (TerraSAR-X and COSMO-Skymed) offers a tremendous improvement in to- mographic reconstruction of urban areas and man-made in- frastructure. The high resolution fits well to the inherent scale of buildings (floor height, distance of windows, etc.). This paper demonstrates the tomographic potential of these SARs and the achievable quality on the basis of TerraSAR-X spotlight dataof urban environment. A new Wiener-type regularization to the singular-value decomposition method—equivalent to a maximum a posteriori estimator—for TomoSAR is introduced and is ex- tended to the differential case (4-D, i.e., space–time). Different model selection schemes for the estimation of the number ofscatterers in a resolution cell are compared and proven to be ap- plicable in practice. Two parametric estimation algorithms of the scatterers’ elevation and their velocities are evaluated. First 3-D and 4-D reconstructions of an entire building complex (including its radar reflectivity) with very high level of detail from spaceborneSARdata by pixelwise TomoSAR are presented.
Abstract: The estimation of deformation parameters using persistent scatterer interferometry (PSI) is limited to single dominant coherent scatterers. As such, it rejects layovers wherein multiplescatterers are interfering in the same range-azimuth resolution cell. Differential synthetic aperture radar (SAR) tomography can improve deformation sampling as it has the ability to resolve layovers by separating the interfering scatterers. In this way, both PSI and tomography inevitably require a means to detect coherent scatterers, i.e., to perform hypothesis testing to decide whether a given candidate scatterer is coherent. This paper reports the application of a detection strategy in the context of “tomography as an add-on to PSI”. As the performance of a detector is typically linked to the statistical description of the underlying mathematical model, we investigate how the statistics of the phase instabilities in the PSI analysis are carried forward to the subsequent tomographic analysis. While phase instabilities in PSI are generally modeled as an additive noise term in the interferometric phase model, their impact in SARtomography manifests as a multiplicative disturbance. The detection strategy proposed in this paper allows extending the same quality considerations as used in the prior PSI processing (in terms of the dispersion of the residual phase) to the subsequent tomographic analysis. In particular, the hypothesis testing for the detection of coherent scatterers is implemented such that the expected probability of false alarm is consistent between PSI and tomography. The investigation is supported with empirical analyses on an interferometric data stack comprising 50 TerraSAR-X acquisitions in stripmap mode, over the city of Barcelona, Spain, from 2007–2012.
Synthetic Aperture Radar (SAR) tomography presents the advantage ofmultiple stable targets detection within same pixel. Fast-sup-GLRT (generalized likelihood ratio test based on support estimation) algorithm proved to be an ideal compromise between detection capabilities and computational complexity. In this work, a multi-look version of this detector which exploits the advantages of Capon estimation is examined. Statistical analysis of estimation and detection processes are conducted to compare the performances of sequential non-linear least-squares (NLLS) search and Capon filtering of projected data for double PS identification. Main objective is to exploit the super-resolution advantages of NLLS method without the risk ofmultiple stable targets classification from the same scattering contribution. For the last desiderate, an additional verification is included within the detection step.
Differential synthetic aperture radar (SAR) tomography allows separation ofmultiple coherent scatterers interfering in the same range-azimuth resolution cell as well as the estima- tion of the deformation parameters of each scatterer. In this way, the spatio-temporal tomographic inversion serves as a means to resolve the layover and simultaneously improve de- formation sampling. Compared to metropolitan regions with several man-made structures, the prevalence of coherent scat- terers in the villages of alpine regions is generally low, while at the same time layovers are widespread due to the rugged- ness of the terrain. Moreover, the drastic height variations in the imaged scene necessitate height-dependent compensation of the atmospheric phase delay variations within the tomo- graphic inversion. This paper addresses these concerns while performing experimentson an interferometric stack compris- ing 33 Cosmo-SkyMed strimap images acquired in the sum- mers between 2008-13 over Matter Valley in the Swiss Alps. The results show improved deformation sampling along the layover-affected mountainside.
airborne bistatic synthetic aperture radar (SAR) experiment, conducted early November 2007, using the German satellite TerraSAR-X as transmitter and the German Aerospace Cen- ter’s (DLR) new airborne radar system F-SAR as receiver. The importance of the experiment resides in both its pioneering char- acter and its potential to serve as a test bed for the validation of nonstationary bistatic acquisitions, novel calibration and synchro- nization algorithms, and advanced imaging techniques. Due to the independent operation of the transmitter and receiver, an accu- rate synchronization procedure was needed during processing to make high-resolution imaging feasible. Precise phase-preserving bistatic focusing can only be achieved if time and phase syn- chronization exist. The synchronization approach, based on the evaluation of the range histories of several reference targets, was verified through a separate analysis of the range and Doppler contributions. After successful synchronization, nonstationary fo- cusing was performed using a bistatic backprojection algorithm. During the campaign, stand-alone TerraSAR-X monostatic as well as interoperated TerraSAR-X/F-SAR bistatic data sets were recorded. As expected, the bistatic image shows a space-variant behavior in spatial resolution and in signal-to-noise ratio. Due to the selected configuration, the bistatic image outperforms its monostatic counterpart in almost the complete imaged scene. A detailed comparison between monostatic and bistatic images is given, illustrating the complementarity of both measurements in terms of backscatter and Doppler information. The results are of fundamental importance for the development of future nonsyn- chronized bistatic SAR systems.
In order to experimentally verify this, a bare surface has been analysed. Such a scatterer represents a point tar- get scenario in the vertical direction, since it is the only contribution. Moreover, in order to detect for the single bounce scattering mechanism, the tomographic reconstruc- tion has been carried out in the Pauli1 component. The to- mographic vertical profile is shown in Figure 5(a) and is obtained by means of the MUSIC algorithm. In order to check the resolution of such a MRA tomographic constel- lation, the same scenario has been reconstructed by means of the Fourier beamforming (Figure 5(b)). The expected
allows to compare the results to a reference, uniformly- sampled data set with an oversampling rate representa- tive of a typical satellite staggered-SAR system, i.e., much lower than the oversampling rate of the F-SARdata set, for which azimuth ambiguities are no longer negligible. Figure 4 shows the focused images obtained for a reference system with constant PRI and a stag- gered-SAR system, where the sequence of Figure 2 is used in combination with BLU interpolation. A relative increase of the intensity in low backscatter areas can be observed for staggered-SAR, but only if data are dis- played using a large log-intensity scale. In a typical sat- ellite scenario, due to the much lower signal-to-noise ratio (SNR), this difference would be hardly noticeable.
Super-resolution imaging via compressed sensing (CS) based spectral estimators has been recently introduced to synthetic aperture radar (SAR) tomography. In the case of partial scat- terers, the mainstream has so far been twofold, in that the tomographic reconstruction is conducted by either working directly with multiple looks and/or polarimetric channels or by exploiting the corresponding single-channel second order statistics. In this paper, we unify these two methodologies in the context of covariance fitting. In essence, we exploit the fact that both vertical structures as well as the unknown polarimetric signatures can be approximated in a low dimen- sional subspace. For this purpose, we make use of a wavelet basis in order to sparsely represent vertical structures. Addi- tionally, we synthesize a data adaptive orthonormal basis that spans the space of polarimetric signatures. Finally, we vali- date this approach by using fully polarimetric L-band data ac- quired by the E-SAR sensor of the German Aerospace Center (DLR).
Transferfunktion, die ein kontinuierliches in ein diskretisiertes Signal umwandelt. Da- bei kommt es unweigerlich zu Informationsverlusten durch die Quantisierung. Außerdem kommt es bei digitalen bildgebenden Verfahren häufig zu Rauschen. Bei der Superresoluti- on als Bildrestorationsverfahren kann im Gegensatz zu Einzelbildverfahren der Verlust an Bildinformationen durch Zurückgreifen auf mehrere Bilder ausgeglichen werden. Bei Ein- zelbildverfahren handelt es sich um ein unterbestimmtes Problem und muss deshalb mit Annahmen über den Intensitätsverlauf arbeiten. Durch die Verwendung mehrerer Bilder können weitere Bedingungen aufgestellt, und das schlecht gestellte Optimierungsproblem verbessert werden.
The generalized time warp method for M = 2 has been applied to the “diamond” substack in Fig. 5 by choosing linear and seasonal motion as the base functions. The left image in Fig. 9 shows the TS-X intensity map of the region of interest (marked by a box in Fig. 6). According to Fig. 6, the center of the subsidence pattern, i.e., the “epicenter,” is located on the right upper part of the intensity map. Therefore, together with the seasonal motion results shown in Fig. 8, we can expect the following: 1) only the building structures suffer from thermal dilation, and 2) the closer to the “epicenter” the point is, the bigger is the linear subsidence.
Abstract— Following the success of the first bistatic spaceborne-airborne experiment between TerraSAR-X and F- SAR carried out in November 2007, DLR has performed a second bistatic experiment in July 2008 with new challenging acquisitions. Furthermore, the existing bistatic processing chain has been updated with two significant improvements: a) clock offset synchronisation is now performed without the use of reference targets, and b) SAR imaging is done using a fast focussing technique. The new SAR imaging algorithm, based on the fast factorised backprojection algorithm, has proved very good focussing qualities while dramatically reducing (up to a factor 100 with respect to direct backprojection) the overall computational load. The new processing chain is tested using the image of the first TerraSAR-X experiment. Results of a dual- pol acquisition performed during the second TerraSAR-X/F-SAR experiment and showing the first dual-pol bistatic spaceborne- airborne images are also presented in this paper.
Different from interferometric SAR satellite formations with two nearly identical satellites, like the TanDEM-X and Tandem-L missions, there is also an increased interest in deploying additional receive-only sensors in formation to already existing traditional single sensor SAR missions. These deputy satellites can be much cheaper due to their re- duced mass and energy requirements, allowing deployment of several entities within one single launch. ESA considered the deployment of a companion satellite to the Argentinean SAOCOM sensor  aimed at enhanced sensing of primar- ily forested areas, whereas the SESAME mission proposal targeted the measurement of polar ice mass variations and glacier flow by means of two C-band receivers in forma- tion with a Sentinel-1 follow-on mission . In addition, Germany’s space agency DLR also considers the additional launch of receive-only satellites for the TerraSAR-X follow- on mission. For safety reasons, all these mission proposals have in common the relatively large distance of deputies with respect to the illuminating master/chief satellite. This re- quirement poses several challenges with respect to the SAR
Future synthetic aperture radar (SAR) systems will incorporate multi-channel Digital Beam-Forming (DBF) capabil- ities and operate in new modes. These SAR instruments offer new opportunities but also challenges for calibration. For example on-board real time channel adjustment is unavoidable, but then the on-board digital signal processing capabilities are also readily available in DBF SAR. In any case, current instrument calibration concepts can not be extrapolated to future multi-channel SAR. Thus a new approach is requires required here. This paper reviews the calibration functionality of state-of-the-art spaceborneSAR and then suggest a calibration concept for future SAR.
Differential synthetic aperture radar interferometry (DIn- SAR) has become a powerful tool to measure deformation phenomena at a large scale. Very high accuracy can be at- tained by exploiting the coherent nature ofSAR systems in order to obtain a precision that is in the order of a frac- tion of the wavelength. DifferentialSAR interferometry using a spaceborne platform is already a quite established technique, since the stable trajectory of the satellite ensures the SAR processor will focus the data without introducing undesired artifacts. Also, the fact that large stacks of im- ages are available, has been of great help to develop the aforementioned techniques. However, the airborne case is almost the opposite. First, there exist only very few dif- ferential data sets with in-situ measurements to be able to validate results. But more importantly, the data processing becomes a challenge itself since it is subject to the limita- tions imposed by motion compensation (MoCo). The fact that the platform does not follow an ideally rectilinear tra- jectory arises several drawbacks that must be considered if accuracy is a priority. However, the advantages an airborne platform offers are quite appealing: flexibility in sense of spatial resolution, used wavelength, and data acquisition. Furthermore, the atmosphere has little or no impact, de- pending on the used wavelength and flight altitude, and the costs of upgrading the hardware of an airborne system are insignificant compared to those necessary to launch a new satellite with improved performances.
The key element for an efficient radiometric cali- bration is a precise antenna model. This allow that most of the antenna in-orbit characterization of the SAR antenna, such as verification of the antenna pat- tern can be shifted from the commissioning phase to pre-launch activities. In particular, the antenna model provides not only the shape of the patterns (as required for the radiometric correction across the scene) but also the gain offsets between different beams and sub- swatch (i. e. in the case of ScanSAR and TOPS). Thus, absolute radiometric calibration can rely on one abso- lute calibration factor valid for all beams and SAR modes. However, a suitable set of beams has to be measured in-flight to verify the antenna model. For this purpose the Institute has derived several specific methods and procedures .
Note that, for MSF and DCRCB, the set of tomograms depicting the wings of the selected edifice (Figures 11 – 13 ) present high ambiguity levels among the PLOS height range from −35 m to 60 m and from −10 m to −20 m, with a pseudo-power up to −3 dB and −5 dB, respectively, in the VV polarization. CS and WISE perform reduction of the ambiguity levels; however, the presence of ambiguities is still significant in some positions, especially for VV, which may lead to false detections. The analysis of all polarizations and the succeeding tomograms helps discriminating these ambiguities. The latter makes more sense for the single-look case, since the spatial mixture of sources is avoided. Moreover, we can also compare the single-look response against the multilook response. By instance, the ambiguity levels for CS and WISE in Figure 11 (HH polarization), among the aforesaid PLOS height ranges, are weaker. Furthermore, these same ambiguities do not appear in the succeeding tomograms in Figures 15 – 17 . Without having a priori information on the ROI, we could infer that the sources within such ranges, for the VV polarization in Figure 13 , are indeed ambiguities. Thus, the PLOS height range can be set accordingly, e.g., from −10 m to 30 m.
SL1MMER, respectively. Moreover, false alarm rate P F is illustrated in Fig. 2b as a function of SNR for M-SL1MMER (red) and SL1MMER (orange), respectively. For the case N = 6, the gain of using multiple snapshots regarding P D and P F is comparable. Elevation estimates are shown in Fig. 3, where each dot depicts sample mean of all estimates, with error bar indicating the corresponding standard deviation. The true elevation profiles would be two line segments, one on off- diagonal related to building façade and the other parallel to the x-axis related to ground. Green dashed lines mark true elevation profiles ±1×CRLB. Missing points suggest that detection rate is below 25%. Overall, the elevation estimation accuracy of SL1MMER approaches CRLB and degrades with decreasing elevation distance. On the other hand, M- SL1MMER offers not only much smaller estimation variance, but also better super-resolution (cf. Fig. 2a). SL1MMER performs in particular much more inferiorly with small N and low SNR. On the contrary, even for the case N = 6, reasonable profiles are reconstructed with M-SL1MMER.
A larger area about 1 2 km around the Las Vegas convention center is chosen as the second test site due to a known subsidence discovered in ( Zhu and Bamler, 2011 ). The same processing proce- dures are applied. The pixel classiﬁcation is shown in Fig. 11 . Blue color (PS) ﬁlls up most of the building structures, while yellow (DS) appears mostly on the road. The park located on the lower half of the image is almost black due to temporal decorrelation of vegeta- tion. In Figs. 12 and 13 , the estimation results of PSs and DSs com- bined are demonstrated. The elevation estimates look consistent and reliable for both PS and DS, except part of the building façades due to the layover problem. The linear deformation rate estimates show a subsidence pattern centered at Las Vegas convention center with a maximum of about 15 mm/year.
a reference DEM, i.e. PSs are localized in 3D. Geocoding of the PSs reveals an interesting effect that, for a high number of buildings, PSs may be found beneath the earth surface. Some buildings are even characterized by patterns of Ghost-PSs. An example is given in Figure 1 showing PSs pertinent to a single building of Berlin. The top view onto the distribution of PSs reveals that the majority of PSs are related to two facades and are located in vertical planes. Likely, the deviation from the vertical plane mainly depends on the limited accuracy of the localization in elevation due to the narrow orbital tube of TerraSAR-X. As the average height of the ground level can be estimated from the DEM used for PSI processing, PSs above and below the ground level can be separated (colored in white and red, respectively). On the right part of Figure 1, the spatial distribution of PSs in height is shown from a side view. As SAR sensors in X-Band like TerraSAR-X do not enable to monitor objects below the earth surface, the Ghost-PSs are related to the limited localization capability of PSI.